IMPACT OF HOMOPHILY ON DIFFUSION DYNAMICS OVER SOCIAL NETWORKS

The purpose of this study is to find out under what conditions homophily reinforce the diffusions over social networks or undermines them. To realize this aim, formal modeling approach is utilized and an Agent-Based Model is constructed. Afterwards, diffusion of a non-sticky innovation is investigated with the experiments having varying homophily levels in a social network with two distinct kinds of agents as the primary control variable. The results show that (i) homophily reinforces itself (ii) looking at the macro-behavior of the diffusion, initial increases in the level of homophily has a positive effect on adopted fraction of the population whereas further increases have a negative impact, and (iii) looking at the micro-behavior of the diffusion, increasing homophily can result in local maxima even the macro trend is decreasing. Connectedness and average degrees interacting with social persuasion are the two explanatory remarks in the course of investigating the impact of homophily. As a by-product, the model is also capable of capturing the segregation dynamics over social networks. Future research involves allowing the adopted innovation to lead to value homophily, exploration of the different diffusion initiation types and adoption heuristics.